We appreciate the interest of Sansonno et al. in our recent article.1 Sansonno et al. correctly state that we observed a reduction in the total number of B cells and in the size of the naïve B cell subset in patients with mixed cryoglobulinemia (MC) compared with hepatitis C virus (HCV)-infected patients without MC and healthy controls. The reduced frequency of naïve B cells in HCV-infected patients with MC was associated with decreased ex vivo expression of the antiapoptotic protein Bcl-2 compared with HCV-infected patients without MC and healthy controls, and with an increased in vitro sensitivity of naïve B cells to apoptosis.
We are somewhat perplexed, however, by the comments that our results are contradictory to many published observations showing an increased number of B cells in MC. The two studies2, 3 cited by Sansonno et al. did not report total B cell counts, and the study by Curry et al.2 focused on CD5-positive B cells. The majority of studies reported frequencies of B cells and B cell subsets rather than absolute counts. Furthermore, a consensus has not been reached regarding alterations in the various B cell subsets in MC. Cacoub and colleagues4 found reductions in the percentage of naïve B cells in MC patients compared with normal controls and, as in our study, the size of the naïve B cell subset was restored after rituximab therapy. Racanelli et al.5 and Oliviero et al.6 found no changes in B cell frequency, whereas Ni et al.7 observed an increase, but only in patients suffering from severe liver disease. In the figure that Sansonno et al. included in their letter, it is unusual that the patients were categorized based on B cell percentages of <5%, 6%-10%, 11%-20%, and >20%. This unusual stratification renders comparisons with other studies difficult. Nonetheless, previous work found no increased prevalence of cryoglobulinemia in patients with B cell frequencies of >20% of the lymphocyte population.7
The decreased B cell numbers in our study do not contradict the large expansions in specific B cell clones observed in cryoglobulinemia. Clonal B cell populations in the activated/memory B cell subset drive MC, and we did not identify reductions in this subset. Rather, we found the percentage of these cells in the total B cell population to be increased and resistant to apoptosis.
With regard to the comment that our observations may be related to changes in B cell homing, we observed differential chemokine receptor expression on B cells from healthy controls and HCV patients, but no significant difference among HCV patients with and without MC. Thus, there is no evidence of differential B cell homing.
The high percentage of men in our study is indeed interesting, and the possibility that sex and perhaps hormonal differences may explain the differences reported among studies is worth exploring. Of note, a large retrospective cohort study of 146,394 patients of Veterans Affairs health care facilities in the United States, in which 97% of the patients were male, also demonstrated an increased risk of cryoglobulinemia for HCV-infected men.8